1. Recommendation, first and unhedged
Open with the verdict, the amount, the instrument and the valuation. A committee reading a memo that withholds the recommendation until page six spends those six pages guessing rather than evaluating.
State confidence alongside the verdict. "Invest, high confidence" and "Invest, low confidence pending customer references" lead to very different conversations, and only one of them is honest when three of your five key inputs are estimates.
2. The opportunity in three sentences
What the company does, who pays for it, and why that changed recently. If the third sentence is missing, the deal is usually a good idea that was also a good idea five years ago — which raises the question of why it has not been built.
3. Market, built bottom-up
Top-down market sizing ("a 1% share of a $40bn market") is the single most common reason a memo fails scrutiny. Build the number from units: how many buyers exist, what proportion are reachable through the stated channel, and what they can plausibly pay.
Label each input. A customer count pulled from a public registry is verified. A conversion rate borrowed from a comparable company is an estimate. A price point that no customer has yet paid is an assumption.
| Layer | What it answers | Common failure |
|---|---|---|
| TAM | Everyone with the problem | Quoting an analyst report for an adjacent category |
| SAM | Those reachable by this model and geography | Ignoring segments the product cannot legally serve |
| SOM | Realistically winnable in the plan period | A round percentage with no derivation |
4. Competition and defensibility
Name real competitors with live products, and name the status quo — usually a spreadsheet, an agency, or doing nothing. A memo claiming no competition is describing a market with no demonstrated demand.
Then answer the harder question: after a funded incumbent notices this, what stops them? Network effects, switching costs, proprietary data, regulatory position and distribution are the durable answers. Being earlier and trying harder is not one.
5. Team
Tie the team to this specific plan rather than to general impressiveness. The relevant question is whether these people have shipped something comparable in complexity, and who is missing for the next eighteen months.
6. Financials and unit economics
- Gross margin at current scale and at plan scale, with the difference explained.
- CAC, payback period and retention by cohort, not blended lifetime averages.
- Burn, runway and the specific milestone this round is meant to reach.
- Downside, base and upside cases, each naming the one assumption that drives it.
7. Risks and what would change the answer
List risks with likelihood and impact, and a mitigation or an explicit acceptance for each. Then add the falsification line: the specific observation that would flip your recommendation. A memo without one is advocacy, not analysis.
8. Open questions and evidence appendix
Every unanswered diligence item goes here with the evidence that would close it. The appendix carries the sources — documents, pages, dates — so a reader can check any figure without re-running the work.